In this video, learn about the purpose of the operation excellence pillar.
- [Instructor] Achieving operational excellence. When we're defining this term using the well architected framework, this means you've developed the ability to run your applications successfully at AWS. And the reason for that success is always monitoring. Only by monitoring will you be able to prove that your application is operating properly in the good times and when it's not operating well, so you can make changes. So we're always wanting to look for ways to improve our existing procedures that we've defined when we're operating at AWS, because there's always going to be something improvement wise provided by AWS. An example, you might be using say the Application Load Balancer, and it works just fine. And then Amazon comes out with a feature that allows you to authenticate using the Application Load Balancer. And maybe this is a better consideration for your application that wasn't there before. Achieving operational excellence means that you've learned from your operational failures and through lessons learned from those failures. And how did you know things failed? Because of monitoring. Monitoring also allows me to anticipate the failure and plan for failover. And the failover could be high availability failover or failing over to another location. And this might be an automated solution. And you might be able to build this into your stack, that everything happens automatically when there's a failure you can solve that without having to rely on manual processes. When you're developing your application and as it runs on a daily basis, testing will have to be performed to identify, for example, a single point of failure. Maybe it just appears over time. How can I solve that potential problem? Can I remove it? Can I mitigate it? Only by getting into this detail monitoring and analyzing my application as it operates, can I move towards operational excellence. The best practices defined by the operational excellence pillar include, first of all, effectively planning for success, planning in a team-like approach, meaning what are your developers doing? What does the business want? What does operations want? Can we work together? Next is operations. How do I design my workload so it can operate for expected outcomes? In order to get expected outcomes, you're going to have to monitor what is running in the cloud to actually get the details back from the monitoring service. For example, Amazon's CloudWatch, or maybe a third party tool that provides you with details as to what's going right and what's going wrong. Finally, we have to evolve because the cloud is constantly evolving. As we learn things from failures, from education, we want to share those lessons that are learned. We want to share that information and knowledge as we move towards that elusive goal, but attainable goal of operational excellence.